1
DATA ANALYSIS: ASSIGNMENT #2 (OPTIONAL)
OPTION A: DESCRIPTIVE STATISTICS
DESCRIPTIVE AND INFERENTIAL STATISTICS
20% OF YOUR FINAL GRADE
Faculty of Social Science and Humanities (FSSH)
WINTER 2021
(PAGE LIMIT: NO MORE THAN 10 PAGES-1 .docx file and 1 .sav file)
DUE DATE: WEEK OF MARCH 29TH IN THE FIRST 10 MIN OF LAB VIA CANVAS
Please start now. These assignments take time.
OPTION 1:
UNIVARIATE ANALYSES & CROSSTABULATIONS: Part A: SPSS Instructions
Complete this assignment using the DEMO(1).sav data set using the variables below (CANVAS). Run appropriate Univariate
Statistics and Select Cases on the following variables:
ABDEFECT: abortion due to strong chance of serious defect
POLVIEWS: think of self as liberal or conservative
CHILDS: # children
1. Use your discretion and run the most appropriate Univariate Statistics on all of the above variables. Accompany your
analysis with appropriate visual representations of the data. HINT: FIGURE OUT LEVELS OF MEASUREMENT
FIRST and run stats accordingly.
2. Recode POLVIEWS into two response attributes, Liberal and Conservative (Moderate can go as system missing).
Run univariates on it…has the level of measurement changed? Why or why not?
3. Manipulate your sample using the “select cases” command. Select only those respondents who are above the age
of 20 and are MALE. Now, re-run a frequency distribution on # of children and see how your sample and sample
size changed from the original variable.
Zero-order Crosstabulation and Elaborations:
Use the “DEMO1.sav data” set to complete this assignment. This data set is available on CANVAS. Also, use lecture/lab
hand out on Cross-Tabs to guide you! Before starting clearly understand what your research question and hypotheses are
and what your IV and DVs are.
You are a researcher who wants to assess the following variables in a bivariate and trivariate relationship(s):
ABSINGLE: respondent is not married, therefore abortion (yes, no)
SEX: respondent’s sex (male, female)
CONTROL FOR ELABORATION: RACE (TWO RESPONSE ATTRIBUTES)
1. Find and examine the distributions of the variables you are using. Decide which is the IV is and which is the
DV. Determine if any recoding is necessary. This process will engage in “logical” recoding of certain
variables. All variables should be recoded so that there are only two response attributes. After
recoding, make sure all variables have proper defined values and value labels.
2
2. Run a 2×2 zero order crosstabulation on the original hypothesis being tested. Check and see if the
relationship is significant. Only if it is significant run the appropriate elaboration.
3. Run 1 elaboration on the ZERO ORDER cross-tab, using RACE. Be sure your control variable has two
response attributes, if not recode accordingly.
4. Interpret your findings using Subheadings and write up your analysis. Make sure your interpretation
discusses the following:
Be sure to address the following for UNIVARIATE STATISTICS:
Purpose of analysis.
Technical and substantive interpretation for all statistics
Describe the nature of your results for each variable distribution by discussing the type of variable and
level of measurement.
Be sure to discuss measures of central tendencies, dispersion, and charts.
Also, discuss skewness/kurtosis where applicable
How do the distributions for the select cases variables change from the original variable’s univariate
stats? What do the univariate results on those particular variables suggest?
Summary of findings.
Submit both written and SPSS outputs as an Appendix.
Be sure to address the following for CROSSTABULATIONS:
-The research purpose/question? You should have a null and alternate hypothesis stated at the
very beginning plus an arrow diagram; Logical discussion of controls and the logic of recoding
(Data Modification Section).
-Cross-tab results with discussion of Conditional distributions, column percentages, Chi-square
(statistical dependence or not?), and Measures of Association (strength of relationship).
-Upon elaboration, has the original relationship modified. What are the observed patterns?
Explain. Why do you think your results are the way they are? Sociologically explain [briefly].
-Do the results support/refute your research hypotheses? Did you expect this particular outcome
or not? Why?
3
SSCI 2910 DATA ANALYSIS-DESCRIPTIVE STATISTICS
Faculty of Social Science and Humanities
WINTER 2021
Mark Breakdown OPTION #1 – Assignment #2 (Univariate Statistics & Crosstabulations-20% of Final Grade)
UNIVARIATE STATISTICS & CROSSTABULATIONS

Total Score (out of 40): ______________
Lateness Deductions (if any, 2 marks/day): _____________
OR DO OPTION #2 BELOW
4
DATA ANALYSIS: ASSIGNMENT #2 (OPTIONAL)
OPTION B: INFERENTIAL STATISTICS
COMPARISON OF MEANS & PREDICTIONS/ASSOCIATIONS
20% OF YOUR FINAL GRADE
Faculty of Social Science and Humanities (FSSH)
WINTER 2021
(PAGE LIMIT: NO MORE THAN 10 PAGES-1 .docx file and 1 .sav file)
DUE DATE: WEEK OF MARCH 29TH IN THE FIRST 10 MIN OF LAB VIA CANVAS
Please start now. These assignments take time.
OPTION #2: COMPARISON OF MEANS: INDEPENDENT SAMPLES T-TEST, ONE-WAY ANOVA OR
CORRELATION & REGRESSION USING THE VARIABLES BELOW
Use the GSS93 run the following inference-based statistics.
MARITAL STATUS (MARITAL)
HIGHEST YEAR OF SCHOOL COMPLETED (EDUC)
LABOUR FORCE STATUS (WRKSTAT)
TV HOURS
PART A: SPSS Instructions
Using the above variables and your own discretion run at least two inference based statistical analyses and interpret
your findings using technical and substantive interpretations.
1. Independent samples t-test OR
2. One Way ANOVA, as well as POST HOC OR
3. CORRELATION & REGRESSION
BEFORE RUNNING THESE TESTS THINK ABOUT ANY RECODING YOU MAY NEED TO DO TO MEET THE CRITERIA OF THE TESTS.
PART B: RESULTS/INTERPRETATION
Use subheadings of each test to organize your write-up.
Independent samples t-test
1. State the research purpose of the test and the null and alternative hypotheses.
2. Discuss all components of the results, including Levene’s Test, technically and substantively.
3. Is there a mean difference between the two groups in question? And if so, is it significant?
4. Is the null hypothesis rejected or do we fail to reject the null… at an alpha=.05? If so, what does this
indicate?
5. Overall, what are your concluding statements (without using #’s)? Link your conclusion to a sociological
explanation.
One Way ANOVA
1. State the research purpose of the test and the null and alternative hypotheses.
2. Discuss all components of the results, both technically and substantive.
3. What is the grand mean for the DV? How does it compare to the group means?
4. Is the null hypothesis rejected or do we fail to reject the null, at an alpha=.05? If so, what does this indicate?
5. Discuss the results of the post-hoc test (Bonferroni).
5
5. Overall, what are your concluding statements (without using #’s)? Link your conclusions to a sociological
explanation.
Correlation & Regression
1. State the research purpose of the test and the null and alternative hypotheses.
2. Discuss all components of the results, both technically and substantive.
3. What do the descriptives tell us, the correlations? Global F test, Adjusted R-square, Casewise Diagnostics?
Linearity, Homoscedasticity, Multicollinearity, VIF, Tolerance? Unstandardized Beta, Standardized Beta, tvalues and significance? Is the null hypothesis rejected or do we fail to reject the null, at an alpha=.05? If so,
what does this indicate?
4. Overall, what are your concluding statements (without using #’s)? Link your conclusions to a sociological
explanation.
5. Make a prediction based on your model using y=a+bx equation
Presentation of Work Does Matter:
Use subheadings to let the reader understand your analysis. It is important to be organized in your statistical writeups.
6
SSCI 2910 DATA ANALYSIS
Faculty of Social Science & Humanities
Winter 2021
Mark Breakdown –COMPARISON OF MEANS-TTEST/ONE WAY ANOVA/REGRESSION-20% of Final Grade)
Student Name____________________________

Total Score out of 50 marks: ___________
Lateness Deductions (if any, 2 marks/day): _______________
COMMENTS: